Upload a strong frame
Use a sharp, well-composed image with clear subject separation. Avoid tiny faces, unreadable products, or details that are already distorted.
Image to video AI Give a still image somewhere to go.
Animate a portrait, product photo, illustration, or AI-generated frame by describing the camera movement, subject action, environmental motion, and mood of the shot.
- Start-frame animation
- Multi-reference options
- Portrait and landscape ratios
Animate a stillKeep the frame. Add the movement.
Let the image define appearance and composition while the prompt directs motion and time.
Slow push in; the subject blinks and turns slightly as soft curtains move behind them…
Seedance 2.5Veo 3.1 FastKling 3
How image-to-video works
Lock the look. Direct the motion.
A start image already defines the subject, light, composition, and style. The motion prompt should focus on what changes through time rather than redescribing the whole picture.
Describe the motion
Combine camera movement, subject action, secondary environmental movement, pace, and a simple ending state.
Choose model and format
Select by reference support and motion style, then match the ratio to the source or the platform where the clip will be used.
Why start from an image
More control over the first frame.
Image-to-video is useful when realism, product shape, character appearance, or an exact composition matters more than complete invention.
Animate portraits and photos
Add subtle expression, posture, hair, fabric, or camera movement without starting the scene from text alone.
Bring product shots to life
Create reveals, light sweeps, turntable-style motion, atmosphere, and social campaign clips from a product frame.
Guide with more references
Where the model supports it, add several images to communicate identity, products, palette, or visual direction.
Control start and end
Supported models can use start and end frames to direct the opening and destination of a generated shot.
Model guide
Pick the image-to-video control you need.
Some models accept one start image, some add an end frame, and others support multiple references. Loura AI changes the composer controls to match the selected provider.
Seedance 2.5
ByteDance's newest cinematic model, up to 30s.
550 credits for an 8-second clip at 720p
Veo 3.1 Fast
Fast Veo 3.1, a great default for iteration.
235 credits for an 8-second clip at 720p
Kling 3
Smooth, controlled camera movement.
295 credits for an 8-second clip
Grok Imagine 1.5
Fast and inexpensive, with sound.
190 credits for an 8-second clip at 480p
Motion prompt examples
Describe motion, not the uploaded pixels.
A practical photo-to-video prompt uses three layers: camera movement, scene action, and atmospheric details.
Browse more prompt ideasPortrait animation
Slow camera push forward. The subject blinks naturally and turns slightly toward the light; hair moves in a gentle breeze. Soft breathing, quiet expression, no sudden facial change.
Product motion
Camera makes a slow half-orbit around the product. A controlled highlight travels across the surface while fine mist drifts in the background. Keep label and geometry stable.
Landscape atmosphere
Slow pan right across the scene. Clouds drift, tall grass moves in layered wind, distant birds cross once, warm light gradually breaks through, calm cinematic pace.
Questions creators ask
Image-to-video questions, answered.
How do I turn an image into a video with AI?
Open Video mode, select a model that supports a start frame, upload the image, describe the motion, choose duration and ratio, then generate.
What should an image-to-video prompt include?
Focus on camera movement, subject action, environmental movement, pace, and atmosphere. The uploaded image already supplies the appearance and opening composition.
Can I animate a real photo?
Yes. Clear portraits, product photos, landscapes, and illustrations can be used as start frames. Review the generated motion carefully when facial identity or exact product details matter.
Can I upload more than one image?
Yes, with models that accept reference images: Grok Imagine 1.5 takes up to 7, the Seedance models up to 5, and Veo 3.1 Fast up to 3. Kling 3 and Veo 3.1 use a start frame and an optional end frame instead.
Why does the generated video change my image?
The model synthesizes new frames rather than moving the original pixels mechanically. Clear references, restrained motion, explicit preservation instructions, and the right model reduce drift.
What aspect ratio should I choose?
Use 9:16 for Shorts, Reels, and TikTok; 16:9 for landscape video; and square or portrait formats for feed placements. Matching the source ratio can reduce unwanted reframing.
Your image already has the look. Give it motion.
Upload a clear frame, describe a restrained movement, and choose the model whose reference controls fit the shot.
Turn an image into video